J. Agrlc. Food Chem. 1001, 39, 1369-1373
1309
Emulsifying Properties of Pure and Mixed asl-and &Casein Fractions: Effects of Chemical Glycosylation Philippe Cayot,' Jean-Luc Courthaudon, and Denis Lorient DBpartement de Biochimie et Toxicologie Alimentaires, Ecole Nationale SupBrieure de Biologie Appliquke A la Nutrition et A YAlimentation, Campus Universitaire de Montmuzard, 21000 Dijon, France a,1- and @-caseinswere prepared from whole casein by batchwise ion-exchange chromatography before being chemically galactosylated. Solubility of these pure fractions was found to be minimal around the isoelectric point (pH 4-5). The glycosylation and the increase of the ionic strength improved the solubility in the range of isoelectric pH and had a small effect on both sides of this pH. Glycosylation had little effect on the emulsifying index. Emulsions made from a,l-caseins do not have the same stability as emulsions from 0-caseins. Nevertheless, glycosylation does not influence the behavior of each fraction as far as emulsion stability is concerned. Binary mixtures of asl-and 8-caseins had a lower emulsifying activity index than the arithmetical calculated value. Adding galactosylated &casein to those binary mixtures improved the emulsifying activity of the resulting ternary mixtures. Then, glycosylation would limit the intermolecular interactions rather than improve the properties of the monomeric casein fractions.
It is possible to explain the good emulsifying properties of caseinates since their primary structure has been determined a,l-casein (Mercier et al., 1971),8-casein(Ribadeau-Dumas et al., 1972), and K-casein (Mercier et al., 1973). Due to their amphiphilic character and the absence of any ordered secondary structure, the principal asl-and &casein fractions have an unfolded structure facilitating their adsorption at the water-oil interface. This relationship has been confirmed by Lee et al. (1987), who pointed out that @-caseinsurface properties were lost by enzymatic hydrolysis of either the very hydrophobic C-terminal fragment (193-209) or the very hydrophilic N-terminal fragment (1-25). The removal of the N-terminal (1-40) or the C-terminal (134-199) regions by a limited proteolysis has been found to alter aS1-caseintensioactive properties (Shimizu et al., 1983). The investigation of the competitive adsorption of individual caseins in oil-in-water emulsions showed that &casein preferentially adsorbs at the oil-water interface when a mixture of asl- and @-caseinsis used (Dickinson et al., 1988); the protein which adsorbs most rapidly remains predominant a t the interface (Brock and Enser, 1987). Despite the importance of kinetic phenomena, 8-casein added to an emulsion made with asl-casein is able to replace it at the interface, whereas the inverse phenomenon is not observed (Dickinson, 1989a,b). There is nevertheless no preferential adsorption at the oil-water interface during homogenization (Robson and Dalgleish, 1987). Casein fractions are efficient at interfaces under monomeric state. However, by associating very easily to each other in aqueous solution as composite complexes, mixed casein fractions often have unforeseeable behaviors; so a mixture of aB1-,8-, and K-caseins in a 41411 ratio has a surface viscosity close to that of entire caseinate, while a 111a,l- and 8-casein mixture has a lower interfacial shear viscosity (Dickinson, 1989a,b). This can lead to the prediction of the preponderant role of K-casein, whose excellent foaming properties are already known (Lorient et al., 1989). This study aims at a better understanding of the role played by the major a81-and @-caseinfractions in emulsions
and at a simulation of the role of K-casein. The latter can be mimed by using chemically galactosylated ael- and &casein fractions since glycosylation was found to improve solubility and viscosity of whole casein (Courthaudon et al., 1989)or to enhance the emulsifyingproperties (between pH 1and 6) of the &lactoglobulin (Bertrand-Harb et al., 1990). In the same way, the emulsifying activities of ethylalkylated 8-casein were higher than that of native &casein, though the solubility was not modified (Chobert et al., 1990). The chemical modification of caseins by covalent attachment of hydrophobic groups did not increase conversely the emulsifyingproperties (Chobert et al., 1987). We looked for the improvement of the emulsifying properties by binding hydrophilic groups like aldoses. MATERIALS AND METHODS Preparation of Whole Caseinate. Bovine casein was obtained from fresh (nonrefrigerated)skimmed milk by isoelectric precipitation at pH 4.6. Then it was washed with deionized water and dissolved at pH 7 by addition of sodium hydroxide. This process of precipitation and resolubilization was repeated twice, and the caseinate solution was eventually freeze-dried. Purification of a.1- and ,%Casein Fractions. Casein fractions were obtained by batchwise chromatography following a method inspired by that of Mercier et al. (1968) as modified by Davis and Law (1977). The separation was made by using strong ion-exchangerQ-SepharoseFast Flow from Pharmacia in a dissociating buffer containing 4 M urea, 6.5 X 10-5 M dithioM Bis-Tris-propanehydrochloride, pH 8, threitol, and 20 X containing 0.01 wt % sodium azide. Each casein fraction was both washed and concentrated by ultrafiltration (MilliporePellicon System equipped with a polysulfone Millipore PTGC membrane with a cutoff point of 10 OOO Da) and then freezedried. The purity of each casein fraction was checked by polyacrylamide gel electrophoresis (7.5% w/v) at pH 8.3 in a 6 M urea and 2-mercaptoethanoldissociating medium according to the method of Maurer (1971). Glycosylation of Casein Fractions. The covalent binding of galactoseto the c-lysyl residues of the a.1- and @-caseins as well as the control of the glycosylation level was made following the methods described by Courthaudon et al. (1989). The level of modification, T , of the a.1- and @-casein(i.e., the percentage of 'hidden" t-lysylresidues)was calculatedaccording to the relationships 0 1991 American Chemical Society
Cayot et al.
1370 J. Agdc. Food Chem., Vol. 39, No. 8, 1991
w
A
B
100B
X
(D
TI C
z .n -3
80-
'
BG
BG
60-
40B
20-
A
AG
-a
02
4
5
a PH
Figure 1. Influence of pH and ionic strength, p = ~ ~ = l ' / 2 c(zi i z=~ ionic valence; ci = ionic concentration) on solubility of a,l-casein (A), galactosylated aal-casein (AG),@-casein(B), and galactosylated @-casein(BG). Conditions: temperature, 20 "C; pH 2,4,5, and 8; p = 0 (solid bars) and 0.1 M (hatched bars); protein concentration, 2.5 g L-l. Values for solubility were set at 100%.
where AR is the absorbance of reference TNP-casein (2,4,6-trinitrophenylcasein) at 420 nm and AGis the absorbance of TNPgalactosylated casein at 420 nm. Solubility Measurement. A 2.5 g L-l protein solution was centrifuged (4000g, 1 h, 20 "C), and the solubility value was obtained from the soluble protein content in the supernatant determined according to the method of Lowry et al. (1951). Emulsifying Properties. The emulsifying properties were assessed with 25% v/v oil-in-water emulsions. The 2.5 g L-l protein solution was mixed with soya oil by using a Polytron PVC 2 homogenizerat 19 500 rpm during 30 s for emulsion activity and at 12 OOO rpm during 2 min for emulsion stability. Theemulsifyingactivitywasdefinedby the emulsifyingactivity index (EAI) of Pearce and Kinsella (1978). To appraise the emulsionstability,we defined an emulsifying stability index (ESI) by t=sO Inin
ESI = 1 / 1t=O
V ( t )dt
where V ( t )is the curve representing the volume of coalesced oil as a function of time, after centrifugation (3800g)for 60 min (one measureper 10 min). The higher the value of ESI, the better the stability. RESULTS AND DISCUSSION
Glycosylation Level of the Pure Casein Fractions. The levels of glycosylation were 69 f 4% with a,l-casein (i.e., 9.7 f 0.6 mol of glucide/mol of protein; MM = 23 900) and 84 f 2 % with @-casein(i.e., 9.2 f 0.2 mol of galactose/ mol of protein; MM = 24 100). Influence of pH and Glycosylationon the Solubility of all-and @-Casein.The highest solubility values were obtained at pH 8 and pH 2, due to a very high charge
density (electrostatic repulsions): a t pH 8, &casein has a net charge of -6 and a,l-casein -15, whereas they have, respectively, +15 and +20 net charge a t pH 2. At pH 5, close to the isoelectric point of asl- (4.96) and @-casein (5.19), the electrostatic attractionsare maximum and cause a lowering of solubility (Figure 1). The solubilizing effect of sodium chloride is only observed a t pH 5 and especially with @-casein.On the contrary, the solubility is decreased by NaCl a t pH 2, while NaCl has very little effect a t pH 8; as a matter of fact, the number of total charges is higher a t pH 8 than a t pH 2. The solubility is increased by glycosylation and NaCl a t pH 5 (especially with a,l-casein) owing to a partial reduction of the electrostatic attractions. An opposite effect is observed a t pH 2 and pH 8 since the charges are lowered by glycosylation. The results are slightly different with whole caseinate. Courthaudon et al. (1989) actually noticed a highly improved solubility a t the isoelectricpoint (pH 4.6) without salt because of the high solubility of glycosylated @-casein. Influence of pH and Glycosylationon Emulsifying Properties of Caseins. It is noteworthy by comparing Figures 1and 2 that the curves of EA1 and solubilityagainst pH are superimposable (standard deviations of Figures 1 and 2 are very low and so are not indicated). The diffusion of proteins to the oil-water interface depends indeed on their solubility. A higher interfacial area can be covered by a,l-casein a t pH 8 than by @-casein (in the same quantities in solution) because of a superior charge density for a81-casein(electrostatic repulsions are more important) and because of a more pronounced amphipolarity for a81casein. The better emulsifying activity of a,l-casein would also be explained by the presence of two hydrophobic "anchorage" sites (residues 1-40 and 139-199) in its primary structure. The effect of glycosylation seems to be minimal a t pH 2 and pH 8 (Figure 2), where the emulsifying activity is already high.
Emulsifying Properties of
cyol-
J. Agric. FoodChem., Vol. 39, No. 8, 1991 1371
and &Caseins
-2,500-
A
P
'? N E U - 2,000-
AG
2
A
1,500-
1,000
500
n v
2
6
5
a
PH
Figure 2. Influence of pH and ionic strength, p, on the emulsifying activity index (EAI) of all-casein (A), galactosylated a,l-casein (AG), &casein (B), galactosylated 8-casein (BG), and whole casein (reference = R). Conditions: temperature, 20 "C; pH 2,4,5, and 8; p = 0 (solid bars) and 0.1 M (hatched bars); protein concentration, 2.5 g L-l. Table I. Comparison of Emulsifying Stability between a,l-Casein (ad,@-Casein(@),Galactosylated a,l-Casein (aoig),and Galactosylated @-Casein(Bg) as Functions of pH and p* ESI,10-2 mL-1 min-1 caseins
2,0001,800-
0 0.1
2.1 6.7
4.0 7.4
40.0 19.0
15.4 15.4
5
0 0.1
0.7 0.9
0.6 2.1
0.6 6.9
0.7 6.3
8
0 0.1
+=J +-
66.7
13.8 6.8
5.0 5.4
2
50.0
Conditions: temperature, 20 "C; protein concentration, 2.5 g
1,6001.4001,200
-I
1,000t-7 0 10
I
I
I
I
I
20
30
40
50
60
L-1.
@-caseingives higher emulsion stabilities than a81-casein at pH 2 (Table I). Since the net positive charges of both @-caseinand a,l-casein are about the same at this pH, this effect could be attributed to enhanced hydrophobic interactions of &casein (segment 193-209 in @-caseinis more hydrophobic than segments 1-40 and 139-199 in a,l-casein). At pH 8, the importance of electrostatic repulsions gives a opposite effect (Table I). Whatever the pH value, the emulsion stability is generally unchanged by the glycosylation. EmulsifyingPropertiesof Binary a81-and @-Casein Mixtures (Figure 3). For possible competitive or synergistic effects to be observed, the conditions of the experimental medium were chosen so that the asl-and @-caseinemulsifying properties were very different. Whatever the relative ratio of two fractions, the emulsifying activity of the mixture at pH 8 was less than the arithmetical average calculated from the emulsifying activities of a81-and &caseins (dotted line; Figure 31). This effect is even more enlarged a t pH 6, where the electrostatic repulsions are lower (Figure 311). This effect could be ascribed to a competition between the casein fractions. Preferential adsorption of @-casein a t the oil-water interface (Dickinson, 1989a,b) would
I
I
I
I
70
80
90
100
Percentage of (I-casein
T m l ,600q-
,
\
1,200
l.loo] 1.000
t 7 1
1
800 0
I
1
1
I
I
I
10
20
30
40
50
60
1
I
70 80 Percentage of
1
1
100 @-casein
90
Figure 3. Emulsifyingactivity index (EAI)of standardmixtures whole casein (*),andthe mixture's of cu,l-caseinand &casein calculated arithmetical EA1 (- - -). Conditions: temperature, 20 "C; p = 0 M; protein concentration, 2.5 g L-l. (I) Plots of EA1 at pH 8; (11) plots of EA1 at pH 6.
(e),
attenuate the influence of a81-casein, although it has a better surface activity (Figure 2) and a better emulsifying stability (Table I) at high pH values. This result could be
Cayot et al.
1372 J. Agric. FoodChem.. Vol. 39,No. 8, 1991
Table 11. Comparison of Emulsifying Activity Index (EAI) between Pure Caseins and Binary and Ternary Mixtures of Caseins. EAI, m2g-l
measd theor pure or mixed constituents 2288 i 133 accasein 1463 68 @-casein 1407 32 galactosylated @-casein 1875 1419 59 ad + @-casein(1/1) 1453 1401 f 121 @ + @g-casein(1/1) 1741 192 1731 a,~ + @ + @g-casein( l / l / l ) 1752 f 139 whole casein a Conditions: temperature, 20 OC; pH 8; p = 0 M; protein concentration, 2.5 g L-l.
*
*
nt'
T
1.500J
T
' 2001 I 1,100
+--r----T---7--r--l---
0
10
20
30
40 50 60 70 80 90 100 Percentage of golactosylatad 6-cassln
Figure 4. Emulsifyingactivity index (EAI)of standard mixtures of @-caseinand galactosylated 8-casein (0)and the mixture's calculated arithmetical EA1 (- - -). Conditions: temperature, 20 OC; pH 8; = 0 M; protein concentration, 2.5 g L-l.
put in relation with the interfacial shear viscosity of a,lcasein 10 times higher than that of @-caseinat pH 7 (Dickinson, 1989a,b). However, it seems likely that association of casein fractions in solution would rather be responsible for the depressing effect of EA1 in anl-and @-caseinmixtures. According to Schmidt (1982), a,l-@-casein heterogeneous associations (with anl-to &casein molar ratios greater than 1) develop preferentially to a,l-casein/a,l-casein or &casein/@-caseinhomogeneous associations. In this way, although a,l-casein is a better surfactant than @-casein, a,l-casein would be more associated in complexes (than @-casein)and therefore less available in monomeric state (the only surface active state) to participate at the emulsion. Emulsifying Properties of Ternary Casein Mixtures. The EA1 of a ternary equimolar mixture of galactosylated 8-casein/a,l-casein/@-casein ( l / l / l )is close to the EA1 arithmetical average of the components (1741 instead of 1731 m2 g-l) and also close to the EA1 of whole caseinate (1752 m2 g-l) (Table 11). This means that the depressing effect is canceled by equimolar proportional addition of galactosylated @-caseinto a,l-casein/@-casein mixture. Galactosylated @-caseinwould lead to a preferential 8-caseinlgalactosylated @-caseinassociation, which would be low if we refer to the small decrease in EA1 (as compared to the arithmetical average) especially for the mixture of 50 96 @-casein/50% galactosylated @-casein (Figure 4). Whole caseinate and ternary mixture of a,l-casein/@casein/galactosylated @-casein( l / l / l )do not behave as a a,l-casein/@-casein(1/1) mixture does. These differences would be likely consequent upon the dissociating role played by galactosylated /3-casein in the ternary mixture. In this mixture, the higher proportion of mon-
omeric molecules could be explained by the destructuring effect of glycosylation-for a high level of modification. According to Colas et al. (1988), glycosylation is indeed known to increase the hydrodynamic volume (actuallythe voluminosity) of the casein fractions (lower Huggins coefficient). The hydrophobic bindings should be therefore reduced. The K-casein undergoes nevertheless aggregation and then associates to about 30 monomers of @-casein(Fox and Mulvihill, 1983). The dissociating role of K-casein would be due especially to a preferential association in the whole caseinate which set free the a81casein to the interface. CONCLUSION The interfacial behavior in oil-in-water emulsions of whole caseinate or binary casein mixtures cannot be predicted from that of native or galactosylated a,l- or @-casein. This makes quite obvious the effect of associations on the interfacial behavior of mixtures and especially if mixtures contain destructured components such as highly chemically glycosylated proteins. Therefore, chemical glycosylation would improve the emulsifyingproperties of whole caseinate by diminishing the a,l-casein/@-casein association rather than improving the emulsifying properties of the casein fractions. Consequently, K-casein plays an essential role, albeit in a minority. Being itself a naturally glycosylated protein, one may wonder whether this effect is due to its natural glycosylation or to its protein structure. LITERATURE CITED Bertrand-Harb, C.; Charier, B.; Dalgalarrondo, M.; Chobert, J-M.; HaertlB, T. Condensation of glycosidic and aromatic structures on amino groups of &lactoglobulin B via reductive alkylation. Solubilityand emulsifyingproperties of the protein derivatives. Lait 1990, 71, 205-215. Brock, C. J.; Enser, M. A model system for studying protein binding to hydrophobic surface in emulsions. J. Sci. Food Agri. 1987, 40, 263-273. Chobert, J.-M.; Bertrand-Harb, C.; Nicolas, M.-G.; Gaertner, H. F.; Puigserver, A. J. Solubility and emulsifying properties of caseins chemicallymodified by covalent attachment of L-methionine and L-valine. J. Agric. Food Chem. 1987, 35, 638644.
Chobert, J.-M.; Touate, A.; Bertrand-Harb, C.; Dalgalarrondo, M.; Nicolas, M.-G.; HaertlB, T. Ethyl-substituted &casein. Study of differences between esterified and reductively alkylated @-caseinderivatives. J.Agric.Food Chem. 199O,38,13211326.
Colas, B.; Gobin, C.; Lorient, D. Viscosity and voluminosity of casein chemically modified by reductive alkylation with reducing sugars. J. Dairy Res. 1988,55, 539-546. Courthaudon, J.-L.; Colas, B.; Lorient, D. Covalent binding of glycosyl residues to bovine casein; effects on solubility and viscosity. J. Agric. Food Chem. 1989, 37, 32-36. Davies, D. T.; Law, A. J. R. An improved method for the quantitative fractionation of casein mixtures using ion exchange chromatography. J. Dairy Res. 1977, 44, 213. Dickinson,E.Food Colloids. An overview. Colloids Surf. 1989a, 42, 191-204.
Dickinson, E. Surface and emulsifying properties of caseins. J. Dairy Res. 1989b, 56, 471-477. Dickinson, E.;Rolfe, S. E.; Dalgleish, D. G. Competitive adsorption of a,l-casein and @-caseinin oil-in-water emulsion. Food Hydrocolloids 1988, 2, 397-405. Fox, P. F.;Mulvihill, D. M. Functional properties of caseins, caseinate and casein co-precipitates. Proc. IDF Symp. 1983, 188-259.
Lee, H. S.; Shimizu, M.; Kaminogawa, S.; Yamauchi, K. Emulsifying properties of peptides obtained from the hydrolysates of @-casein.Agric. Biol. Chem. 1987, 51, 161-166.
Emulsifying RoperHes of all-and @-Caseins
Lorient, D.; Close, B.; Courthaudon, J.-L. Surface properties of the bovinecasein componente,relationships betweenstructure and foaming properties. J. Dairy Res. 1989,56, 495-502. Lowry, 0. H.;Rosenbrough, N. J.; Fan, A. L.; Randall, R. J. Protein measurement with Folin reagent. J.Biol. Chem. 1961, 193,265-268.
Maurer, H.R. Disc electrophoresis; de Gruyter: Berlin, 1971. Mercier, J-C.; Maubois, J-L.; Poznanski, S.; Ribadeau-Dumas, B. Bull. SOC.Chim. Biol. 1968,50, 521-530. Mercier, J-C.; Grosclaude, F.; Ribadeau-Dumas, B. Eur. J.Biochem. 1971,23,41-51. Mercier, J-C.; Brignon, G.; Ribadeau-Dumas, B. Eur. J. Biochem. 1973,35,222-235. Pearce, N . K.; Kinsella,J. E. Emulsifyingproperties of proteins: evaluation of turbidimetric technique. J.Agric. Food Chem. 1978,21, 716-723.
J. Ag&.
Food Chem., Vol. 39,No. 8, 1991
1979
Ribadeau-Dumas, B.; Brignon, G.; Grosclaude, F.; Mercier, J-C. Eur. J. Biochem. 1972,25,506514. Robson,E.; Dalgleish,D. Interfacial compositionof sodiumcaseinate emulsions. J. Food Sci. 1987,52,1694-1698. Schmidt,D. G. Associationof caseinsand caseinmicellestructure. In Developments in Dairy Chemistry; Applied Science Publishers: London, 1982; Vol. 1, pp 61-82. Shimizu, D. G.; Takahashi, T.; Kaminogawa, J.; Yamauchi, K. Adsorption onto oil surface and emulsifying properties of bovine a,l-casein in relation to ita molecular structure. J.Agric. Food Chem. 1983,31, 1214-1218. Received for review December 3,1990. Accepted April 2,1991.